Founding piece
Why WilliamFrank
An operating system does three things for the person using it.
It keeps one place where state lives. It lets programs talk to each other, because they read and write that same state. And it means the user never has to know which program is running.
Wealth management has none of the three.
Two questions
There is a test for this, and any advisor can run it against their own firm.
Ask how many systems a household lives in. Then ask what gets re-entered at each step.
Count the custodian, the portfolio accounting, the planning software, the CRM, the document vault, the billing system, the risk tool, the client portal and the e-signature vendor. Most firms stop counting before they finish the list.
The second answer is the one that matters. Every place a name, an address, a beneficiary or a risk tolerance is typed a second time is a place where the record forked. From that moment the firm holds two versions of the same household and no reliable way to know which one is right.
Those two answers describe the next decade of the relationship.
Why the copies exist
The copies are not an accident of procurement. They are what the rails require.
Positions, transactions and balances arrive in files, mostly overnight. Every application that needs the household has to hold its own version, because there is nothing live to read from. The planning software cannot ask the custodian what is true right now, so it keeps a copy and refreshes it when the file lands.
A copy that arrives overnight cannot be shared state. It is a snapshot every system agrees to treat as current until tomorrow's file replaces it.
That is why wealth management feels like a stack of applications rather than a system. The integrations exist, and there are a lot of them. Integration between systems that each hold their own copy is a reconciliation problem, and reconciliation is not the same thing as sharing.
The complexity tax, applied to the stack
I have written for years about complexity as a tax. Cronyism, regulation and bureaucratic layering are one problem wearing different clothes, and it is a regressive tax: it rewards whoever can afford the lawyers and the compliance departments and punishes the small operator who cannot.
The advisor's technology stack is the same tax paid in software. Each tool arrived solving a narrow problem, and then stayed. And billed. Somebody administers it. Somebody reconciles its output. Somebody keeps its assumptions current. There are way too many layers between an advisor and the family they serve, and all of those layers are built on yesterday's technology.
The bill is paid in the one currency an advisor cannot make more of, which is the week.
The elite advisor's choice
Those who have gained a true expertise in solving complex financial problems have faced a difficult choice when growing. Build around people, or keep ownership of tracking the facts. Both work, but neither optimizes the productivity of the elite advisor.
The main issue with running a bunch of software is that complex financial households have complex facts, and those facts change over time. If you do not have a single place that intelligently tracks these facts and is the source of truth, it creates ripple effects to keep facts accurate. Ultimately it forces an elite advisor to spend time tracking facts or spend time managing people.
An AI-fueled single source of accurate facts, driving all the systems, solves for both directions.
What changed
For most of the industry's history there was nothing to do about this except buy another tool.
Agents change what you are buying. Everything before was a tool, and a tool needs an operator. An agent does the work itself, against a written methodology, and shows what it did. Put one on nine copies of a household and it can only automate the re-entry. Put one on a single record the household owns, live rather than nightly, and it can do the work that does not need judgment and organize the work that does.
That is the reason to want an operating system now, given the current trajectory of AI's capability. The rails set the ceiling on what any agent can reach. Buying and bolting on an agent regardless of quality will not likely yield strong results when bolted onto antiquated rails.
What WilliamFrank is
WilliamFrank is one integrated set of agentic rails for a wealth team. Mail, calendar, messages, projects and tasks, people and households, files — with agents in every surface rather than bolted on as a chat window.
A few rules hold it together. Every person has their own agent. Agents propose; people accept. Every write carries a named actor and lands in a record that is never deleted, only superseded. Copper marks anything an agent did. Blue marks a person. One composer sits at the bottom of every screen, because I do not want a hunt-around-for-buttons experience.
All of that is one decision, made once: where state lives, and who is allowed to change it.
Why in the open
I am building this with a small team and a great deal of AI, and I am writing it down as I go.
Partly that is the argument in practice. The essays here are authored by me with WilliamFrank, the agent I built to think and work alongside me. The democratization of intelligence is something I use, in public, on the hardest problem I know.
Partly it is because the first user is us. The next step is to move our own projects, tasks and correspondence onto the platform and off the separate tools we use today. People each running a separate AI, whose outputs the others then feed into their own, is the inefficiency the product exists to remove. The way forward is not copying AI outputs and sharing them back and forth. It's a single, purpose-built AI like what we are building at WilliamFrank.
The founding consortium
Getting off of batch will determine how much value can be accrued with AI, and the operating system is just the orientation to solving for the re-imagined digital user experience — where state lives, before what the screen looks like.
The incumbents in this industry will fall in the next five years. They run on rails that have proven hard to rebuild from inside the firm that was built on them, and every agent they buy inherits those rails.
I would like to invite a small group of advisors to join me in a founding consortium and build what ought to exist in this industry. Advisors who run the work, who want the administration off their desk badly enough to work in something unfinished and say what is wrong with it. You keep your practice and its enterprise value. You get your own agent, your mail and calendar connected the day you join, a household ledger with a source on every fact, and me in the loop every week.
If that is you, the application is on this site.